1.1 CEDAR

1.1 CEDAR 1.1.1 CEDAR Detector Requirements The disadvantage of high energy protons used by NA62 and, consequently, of a high energy secondary beam, is that the kaons cannot be efficiently separated from pions and protons at the beam level. The consequence is that the upstream detectors which measure the momentum and the direction of the kaons are exposed to a particle flux about 17 times larger than the useful (kaon) one. A critical aspect is therefore to positively identify the minority particles of interest, kaons, in a high rate environment before their decay. This will be achieved by placing in the incoming beam a differential Cerenkov counter, CEDAR, filled with hydrogen gas. A FLUKA simulation was used to study the interactions of pions, kaons and protons with the residual gas in the vacuum decay tank and the probability that such an interaction can cause fake triggers was computed. The conclusion is that (in the absence of kaon tagging) the vacuum should be better that 6 10-8 mbar to keep the background to less than one fake event per year. This very challenging requirement can be relaxed by at least an order of magnitude by positively tagging the kaons by means of a CEDAR Cerenkov counter in the beam line, filled with hydrogen gas at an absolute pressure of about 3.6 bar. A necessary part of this kaon identification is the precise timing of the different components in order to guarantee a good rejection of the background due to the accidental overlap of events in the detector. An upgraded form of the CEDAR built for the SPS secondary beams (CERN Report CERN-82-13) will be used, and will be insensitive to pions and protons with minimal accidental mis-tagging. The choice of the Hydrogen gas is dictated by the need to minimise material on the beam line, and hence reduce multiple Coulomb scattering – see Table 1. The foreseen thicknesses of the upstream and downstream aluminium vacuum windows are 150 µm and 200 µm, correspondingly. Table 1: the thickness expressed in X 0 for different gases used in the CEDAR counter. The pressure values are chosen to have the same index of refraction. Helium

Nitrogen

Hydrogen

10

1.7

3

2x400

2x100

150+200

Al window thickness [10-3 X 0 ]

9.0

2.2

3.9

Gas thickness [10-3 X 0 ]

12

35

3.2

Total thickness [10-3 X 0 ]

21

37

6.6

Pressure [bar] Al window thickness [µm]

1

NA62 TD Document

1.1.2 Overview of the CEDAR operation The CERN CEDAR counter (see Figure 1) has been designed to identify particles of a specific mass by making the detector blind to the Cerenkov light produced by particles of different mass. For a given beam momentum, the Cerenkov angle of the light emitted by a particle traversing a gas of a given pressure is a unique function of the mass of the particle and the wavelength of the emitted light. The Cerenkov light emitted by particles of different mass is then not transported by the CEDAR optics through the diaphragm slit onto the light detectors but absorbed on its way, as explained in more detail below. Table 2: Characteristics and resolutions of the CEDAR

Parameter Gas type: n-1 Nominal pressure for kaons θK

Value H2 ~142 x 10-6 3.85 bar 30.9 mrad

Kaon rate

50 MHz

Time resolution

5000 km). The 8 optical windows are made of quartz cylinders of 45 mm diameter and 10 mm thickness. Their ring frame is made of stainless steel and the binding is obtained by heating the rings at 150o C. A layer of Silver was deposited on the circumference of the quartz disk before heating. After cooldown the silver layer is compressed and produces a perfect seal. In a destructive test, a pressure of 300 bar was reached. All optical surfaces are coated with a quarter-wavelength thickness of MgF 2 for minimum reflection at 300 nm wavelength. The diaphragm comprises a disk with 8 elongated apertures and 8 outer and inner segments moved by right-left screw drives. The segments are bolted onto high-precision guided chariots. The 8 screw drives are provided with gears and are turned simultaneously by a gear mounted on ball bearings in a V-groove on the periphery of the disc. The opening can be varied between 0.03 and 20 mm in steps of 0.01 mm with a motor located external to the vessel. The circularity of the aperture shows radial deviations below 0.02 mm. The azimuthal opening of the diaphragm is 8 times 42.6 degrees and covers 95% of the ring.

1.1.4 Adapting the CEDAR for NA62 NA62 will use the West CEDAR filled with hydrogen at up to 5 bar (note: up to 5 bar for pressure scans is needed to see full proton peak). The nose will be shorten from 1243 mm to about 600 mm. The design of the mechanics and cooling system must accommodate the different potential choices of the optical readout system, minimise the flux of muons and neutrons hitting the photo-detectors, and ensure high levels of optical efficiency and stability, as well as minimising temperature gradients in the vicinity of the CEDAR. New photodetectors and associate electronics are needed in order to operate at the required kaon rate. A detailed description of the necessary changes and adaptations is given in the following sections.

1.1.5 Light transport system The optics of the CEDAR light collection must be re-designed to match the photodetectors necessary to handle the high beam flux in NA62. Reflecting the light through 90 degrees will also be necessary to allow the greater space required for photodetectors and preamplifiers and to locate them so as to 6

1.1 CEDAR minimise damage from radiation. The final design of the light collection system is highly correlated with the technology choice for the photodetectors, with different rate, noise limitations and packing density. As a consequence the light may have to be focussed or de-focussed onto the photodetector planes using ellipsoidal mirrors as illustrated in Figure 3 and a lensing system to focus the light onto the active area of the photodetectors.

Figure 3: The Simulation of 100 events in the CEDAR nose cone. The ray tracing of the photons from (right to left) the quartz window (blue), the ellipsoidal mirrors (red), the light collection cones (red) and photodetector plane (green) is illustrated. The optical components are schematic and do not reflect a final geometrical arrangement. The photons are shown in fawn for those that are detected directly and blue if detected after reflection from the cone. Photons that are lost at reflection in the cone are shown in green whilst those photons lost by reflection at the PMT window are indicated in red.

One possible solution, adopted by the NA62 RICH detector, would be the use of light-collection cones. Again, the final design of such light collectors is dependent on the choice of photodetector technology. A proof of principle of the full optical system has to be developed for each of the photodetector options. Initial simulation studies have been undertaken to understand the design of the optical components. These preliminary studies indicate that the depth and design of the lightcollection cones will be critical. A cone design that is too shallow will result in loss of photons at the phototube window due to Fresnel reflection and will suffer from light being reflected back out of the cone. With a cone that is too deep difficulties could arise in the manufacture process to ensure a 7

NA62 TD Document good reflective coating. The preliminary studies indicate that a photon collection efficiency of ~90% should be achievable. This design study assumed ellipsoidal collection cones, with a depth between 10-25mm; a conical design is currently under investigation. The simulation of 100 events overlaid is shown in Figure 3, the photons are generated via a full simulation of the CEDAR.

The mirrors will be fabricated from glass with radii of curvature of the ellipsoidal sections still to be optimised. The light collection system will be a bespoke solution with two approaches under consideration: machining the “cones” out of a solid block, or creating a mandrel and manufacturing the cones out of plastic, glass or carbon fibre. Further studies are in progress to ensure that the light losses in the cones due to reflections and non-normal incidence are kept to a minimum, acceptable level.

1.1.6 Mechanical Support, Cooling and Safety Considerations The major mechanical considerations involved in modifying the Frontend of the CERN West CEDAR for use by NA62 are as follows: • • •

The photodetectors and optical components must be redesigned and located to maximise the capture of Cerenkov light, and mechanical rigidity must ensure optical stability. The preamplifiers and signal-shaping electronics must be located adjacent to the photodetectors with signal-processing electronics removed from hazardous radiation. The existing protective and thermally-insulated metal cover around the nose section of the CEDAR must be replaced. Cooling and insulation must be designed to minimise temperature gradients and thermal instability that might result in unacceptable local variations in the refractive index of the hydrogen gas.

The safety requirements resulting from the necessary mechanical modifications and the use of hydrogen gas are as follows: •

•

8

A nitrogen blanket around the optical-readout electronics and HV is required to eliminate any possibility of an explosion in the event of a hydrogen leak from the CEDAR. The mechanical design will incorporate sensors to monitor the flow of nitrogen and the temperature of the enclosure, as discussed below in the gas and safety sections. The CEDAR must be connected to the vacuum beam pipe at both ends in such a way that a hydrogen leak is not accompanied by any admixture of air in order to prevent any risk of explosion. An important secondary consideration is that damage to sensitive detectors must be minimised by mitigating the effects of hydrogen leaks and the shock wave resulting from the potential rupture of the window at either end of the CEDAR. The proposed way of doing this involves a single aluminium window between the high-pressure hydrogen and vacuum beampipe at each end of the CEDAR with a large vacuum volume to capture any escaping hydrogen. This task is being studied by the Beam group and is now completely separate from any modifications to the CEDAR nose. A preliminary schematic drawing of the pressure protection volume is shown in Figure 5.

1.1 CEDAR

Figure 4: Conceptual Design of the Support Structure and Nitrogen Enclosure.

Figure 5: Conceptual drawing of the retention vacuum volume attached to the upstream end of the CEDAR( in blue). The aluminum window on the upstream side has thickness of 150 µm, the downstream window has a thickness of 200 µm.

1.1.6.1 Adapting the existing CEDAR nose

The proposed design involves replacing the current photo-detection system of 8 PMT’s and the protective and thermally-insulated metal cover forming the front end of the CEDAR. All optical components internal to the CEDAR are already optimised for use with hydrogen, while the seals on the quartz windows are perfectly safe for use with hydrogen at 5 bar.

9

NA62 TD Document Because of their increased size and unavoidable dead space the 8 new sets of photo-detectors will need to be relocated away from the quartz windows in the nose, and light guides will be required to focus the Cerenkov light onto their active surfaces. The mechanical challenge is to preserve the current optical stability and ensure that all the Cerenkov light reaches the photo-detectors. This requires the precision mounting of the photo-detectors and optical components in a rigid, lightweight structure that can be precisely located relative to the quartz windows of the existing nose. A disc-shaped lattice support structure housing the new photodetectors, optics, readout electronics and cooling will be cantilevered off a support cylinder bolted onto the CEDAR flange. This structure will support all cabling and cooling pipes and be enclosed in a gas-tight environmental chamber through which Nitrogen gas will be circulated. A separate protective metal cover [not shown] will enclose the environmental chamber and clip onto the metal cover enclosing the main body of the CEDAR. In addition to removing the heat from the electronics it is important to maintain a constant temperature for all components of the CEDAR nose that are in thermal contact with the hydrogen gas in order to prevent local variations in density that will affect the refractive index. This will be done by a combination of heat removal, using chilled de-ionised water, and thermal insulation. Finally, it is important that the design ensures rapid access to, and replacement of, any PMT or electronic component that may fail during use.

1.1.6.2 The Photodetector Support Structure

Our conceptual design locates the 8 sets of photo-detectors and readout electronics into 8 pods within a lightweight, disc-shaped latticework support structure at a radius of about 30 cm from the beam. Studies indicate that the high-intensity neutron background radiation is rather uniform and that it will not be necessary to vary the radius or azimuthal angle of any of the 8 pods. A customised, ellipsoidal 45o mirror to reflect the Cerenkov light from each of the 8 CEDAR quartz windows towards a set of photo-detectors will be located at the same radial position as the quartz window. A set of light guides, using either internal reflection from Plexiglas (male) cones or reflection from the surfaces of polished metal (female) cones, will channel the light onto the active surfaces of the photodetectors. The cones will be designed with variable light-collection areas to ensure approximate equalisation of the light intensity falling onto each PMT. Provision will be made for fine adjustments to the position and angles of the mirrors. The photodetector support structure will be built in two halves that can be ‘clam-shelled’ around the beampipe and precisely located onto the support cylinder to facilitate installation. Figure 6 shows a cross-section of one half of the support structure with the envisaged layout of optical and electronic components and cooling plates. The cooling pipes and cabling [not shown] will be led around the outside of the structure and out of the environmental chamber to the floor, where appropriate patch panels are situated. Both the support cylinder and support structure will remain fixed in place after installation and the optics and electronics in each of the 8 segments will be mounted in drawers that slide forwards to enable access to the relevant components through the environmental chamber. We envisage a separate, nitrogen-flushed, environmental chamber, made in two sections from lightweight carbon fibre, which surrounds the support structure and is supported off it. Simple non10

1.1 CEDAR load-bearing seals around the beam pipe and support cylinder will ensure the necessary degree of gas-tightness. Ports will be incorporated to enable rapid nitrogen flushing to remove air from the system and to relieve excessive overpressure. The front face of the chamber will incorporate removable panels to enable rapid access to the electronics, cooling and optics within each sector. This chamber will also ensure that no extraneous light enters the optical system and will keep out any dust. The whole structure will be covered by a protective metal casing rigidly clipped to that of the main CEDAR. This casing will have front doors that can be removed or opened to enable unrestricted access to the front of the environmental chamber and hence to all components within the support structure. The thermal environment must be controlled to ensure that the hydrogen gas in all parts of the CEDAR remains at a constant temperature and that any changes in temperature occur only very slowly and without local fluctuations. To this end the insulation within the protective cover surrounding the main body of the CEDAR will be upgraded with better performing, fire-retarding material. Thermal design of the new CEDAR Frontend will concentrate on removing the heat generated by the electronics and insulating those surfaces in thermal contact with hydrogen gas or with the main body of the CEDAR. The three regions that require particular attention are: i) the hydrogen-filled beampipe; ii) the curved surface and front face of the nose in which the quartz windows are situated; and iii) the flange onto which the support cylinder is bolted. Purpose-designed insulation will be incorporated into these different areas, as indicated in yellow in Figure 4. Details of the proposed temperature control and monitoring are given in section 1.1.6.3.

Figure 6 : Cross-sectional Layout of the Support Structure and Nitrogen Enclosure

An installation frame will be required to hold the two halves of the support structure as they are populated with optical, electronic and cooling components in the laboratory and to facilitate system tests prior to installation on beamline. It is likely that this frame will also be useful in supporting the two halves of the structure while they are being bolted to the support cylinder after being craned 11

NA62 TD Document onto the beamline and in facilitating the location of the environmental chamber. It is envisaged that the outer protective metal casing will be craned directly into place.

1.1.6.3 Safety Considerations, Cooling and Monitoring.

All custom made electrical equipment on the detector (photodetectors and electronics) will be enclosed within the environmental chamber that is filled with dry nitrogen gas to prevent the development of an explosive atmosphere in the event of hydrogen leaks. A small overpressure will ensure that a nitrogen atmosphere also surrounds the optical components in the support structure and prevents any possibility of moisture condensing from the atmosphere. If the nitrogen pressure drops, the power will be cut to the photodetectors and readout electronics to prevent the occurrence of any sparks within the enclosure. The heat load from the photodetectors and front-end electronics is not expected to exceed 10 W per sector, or 100 W in total. This heat will be removed as near to source as possible by metal cooling plates, through which chilled de-ionised water flows, in thermal contact with heat sinks on the electronics PCBs. The temperature and flow of water through the cooling manifold will be controlled by electronics monitoring the temperature difference between the hydrogen beampipe close to the nose and that of the main body of the CEDAR and also the rate of change of temperature within the gas enclosure. The aim will be to maintain a stable temperature at the beampipe within ±0.1oC. It is important to monitor the long-term performance of the 45o mirrors and Light Guides in case any of the surfaces should deteriorate. Early warning would enable replacement parts to be fabricated in good time. A system of 8 ultra-violet LEDs will be installed on the CEDAR nose symmetrically spaced between the 8 quartz windows. By flashing the LEDs and monitoring the output from the 8 groups of photo-detectors the overall response of the optical and electronic systems may be monitored. Factoring out the response of the electronics and photodetectors, obtained from separate monitoring, will provide a long-term record of the optical performance of all parts of the system to enable early warning of any deterioration of the optical components. Careful design is necessary to enable separation of the different contributions from the mirrors and light guides, requiring more than one signal from each group of PMT’s. All custom-designed electronics will require certification by CERN for flammable safety and all commercial sensors will need an ATEX certificate for zone 2 (non-sparking) equipment.

1.1.7 Gas system Hydrogen gas has been identified as the most suitable radiator gas to operate the CEDAR detector. Physics performance arguments reveal two basic motivations for this choice: • •

12

The use of Hydrogen limits the radiation length of the gas to 0.2 % of X 0 (instead of 3% for Nitrogen or 1% for Helium). The refractive index, its wavelength dependence and the optical properties of the gas must be suitable to provide 95% efficiency for the identification of 75 GeV/c momentum kaons.

1.1 CEDAR This gas is very flammable and must not mix with Oxygen (or Air) in any place of the system. All zones exposed to the gas system will be declared as flammable gas areas (Zone 2) and all equipment used in these areas must follow the CERN flammable gas safety rules. All exhaust gases lines must be connected to the gas extraction system in TCC8.

1.1.7.1 Outline of the Gas System Design

The Hydrogen gas supply is located in the gas surface building 920, all other gas system components are situated in the experimental area (TCC8) close to the CEDAR detector itself. The relevant detector and gas-system parameters are summarized in Table 4. Table 4: Detector and gas system parameter Radiator gas volume

1.1 m3

Operating gas

100% H 2

Operating pressure

Between 2.5 and 3.5 bara 1

Pressure scans

Vary pressure in small steps around peak

Smallest pressure step during pressure scan

5 mbar

Absolute pressure (density) accuracy